Thursday, September 30, 2010

F300EXR – Review Part 16 – SN mode versus DR Mode

The question came up … is SN mode better than DR mode on the F300EXR? I have always recommended ignoring SN mode for EXR cameras because there was not improvement in noise and there was a definite issue with DR100, which SN forces. I don’t like DR100 because it risks blocked shadows and blown highlights and I see no reason whatsoever to do that.

The setup for this test is simple … camera on a tripod with IS off to get exact framing match and to get crystal sharpness. That way we can compare directly the edges for NR degradation and the flat areas for excess noise. Also, there is sufficient dust to compare smoothing effects of NR. Note that exposures are identical, which helps us see what the tone curves are doing.

Here is the DR400 shot … in A mode at ISO 100 …

DSCF0097_DR400[1]

And here is the SN shot … in EXR SN mode at ISO 100 …

DSCF0098_SN[1]

You should be able to see the darker tones on the flash bodies … especially apparent on the mody of the Nikon flash. The Nikon’s diffuser and the paper to the right of the Pentax flash are much brighter in the second image. So the tone curve difference is very true.

Now, we examine the crops. You will absolutely have to click through to the larger version if you want to see anything in these crops …

SNvsDR_crops[1]

So what do I see? Let’s go left to right through the five crops:

  1. Noise about the same. Detail about the same. A tie.
  2. A tie on noise and detail. I prefer the tone of the EXR DR because I process my images. You may feel the opposite.
  3. There are bits (almost chunks) of dust that are rendered clearly on the DR image and wiped out on the SN image. This is on the left and right underneath the flash. Details on the lens at the front of the flash are also softer on the SN image. This one goes to the DR image.
  4. The lettering seems darker and the edges are more defined on the numbers and last letter in the SN mode, but the edges are more realistic on the AF letters in the DR mode shot, The DR mode shot has better, more 3D rendering of the flash contacts. Let’s call it a draw.
  5. Again, the SN image shows damage drom excess noise reduction. The hairs along the edge are criso in the DR shot and smoother in the SN shot.

I find that the DR mode eeks out a slim victory … but frankly, both work fine. The choice you make depends on you preferred tone curve for the capture.

F300EXR – Review Part 15 – Pretty Sunset Tonight

Ok … this isn’t a review in any real sense. I just caught a pretty nice sunset this evening and thought it worthy of displaying as testament to the Fuji’s ability to capture some kick-ass color …

Remember that I generally recommend that you shoot DR400 on Provia so that the tone curve is as forgiving as possible. That lets you do anything you want later. If you choose to shoot Velvia when the sun is around or there is high contrast for another reason, you may very well regret it as your shadows go black and your highlights blow out. You really want to get all the tones and then spread them around yourself.

So this means that I am arguing for post processing. This is the necessary second have of the creation process. If you prefer to let the cam do everything, then experiment with tone curves like Velvia and Astia, but be aware of the risks.

I picked the youngest up at the bus station this evening and stopped in the parking lot to grab this image of a sunset. The 360mm lens certainly makes it a big disc.

Here is that shot straight from the camera:

DSCF0082_OOC[1] That’s not half bad … but it is a tad weak and the trees need straightening. Might as well saturate it a bit more while I am in there … oh yeah, darken the silhouette as well …

DSCF0082_sunset[1]

Your call … but I’m picking the second one.

We made our way to the post office where I had to send something to my favorite federal government department (the tax man) and something to my son in the UK. When we left the building I became enamored with the beautiful cloud patterns and colors. Wow …

DSCF0083_OOC[1] Again … pretty. Because I shoot DR400, there is detail in the shadows and the bright spot where the sun is just below the horizon is not blown out. This leaves us an opportunity to crank things up to 11 … (reference: This is Spinal Tap.)

DSCF0083_sunset[1] Yes … that’s more like it.

Being color deficient means that I need a rather strong dose to be satisfied with naturally colorful events like sunsets. This output suits me …

Wednesday, September 29, 2010

Dell loses a customer …

Dell really made a bang back some years ago with their fantastic build to order strategy. You could call them up and they would ship you a computer in no time at all.

So when I was looking for a new laptop for work, I looked at Dell first. I also looked at other brands, finding a few that I liked, but Dell’s site is generally the easiest to use (although their user interface fails a lot on various browsers.)

I really like using a small laptop that I plug into a large monitor at home and at the office. 12” is my favorite size, but 13 is also excellent. I am looking for a desktop replacement with 8GB RAM and a fast processor on the order of an i7 620M up through an i7 840QM … these things are really fast and they support hyper-threading for 4 or 8 threads of execution respectively.

I narrowed it down to two laptops … the Alienware M11x and the Studio 14. The studio is a bit bigger than I wanted, but it has the wicked fast i7 840QM … unbelievably quick for a notebook processor. The M11x is nicer in the hand and has every single feature I want except one … the processor is a bit of a joke. It is an i7 640UM. The U stands for “Ultra Low Voltage” but it might just as well stand for “Ultra Slow” … too bad, because I would love to use one of those machines.

Anyway … I bit the bullet and ordered the Dell 10 days ago or thereabouts. I was immediately annoyed by the projected ship data moving out from 24 September to 18 October. Sheesh … a month to build a laptop from parts? Seriously? But I swallowed my irritation because I knew how great the machine would be in day to day use.

I had ordered an extra power supply (always need two) and a nice mouse and mini surge protector … the last two being on sale. The power supply shipped separately and so did the other two as one other shipment. Both arrived several days ago.

It had been a while since I checked the status so I looked yesterday morning and found that it had been cancelled. That seemed really peculiar, as I did not cancel it. I was busy until the afternoon, when I finally got a chance to call. It took several attempts because the commerce site is in the US and calling those numbers connects you with people who cannot see Canadian customers or order numbers. Sheesh …

They finally transferred me to someone in Canada and we discussed the issue. It seems that the reason why a product gets cancelled is a build configuration problem. This is really strange, since cancelling it loses my place in line, and we are already about 1/3 of the way through a month-long build cycle. So pausing it and calling makes way more sense.

I was so annoyed with the thought of getting back in line that I told them to leave it cancelled … that I was done with Dell (and I am.) So I got an RMA for the power supply, which I have packaged and sitting in my car waiting to go to the UPS store (which keeps banker’s hours for reasons unknown.)

The cherry on top of this travesty is a call I got this morning from Dell, informing me that my order had been cancelled … they had actually left a message on my cell yesterday, but I did not get it until well after the issue was terminated.

He went on to tell me that the graphics card I had selected was not available on the Studio 14 (so why the f*ck was I allowed to select it?) and that all we had to do was to redo the configuration and reorder the laptop. And for my trouble, they would give me an extra $20 off over the $30 I was saving by not getting a good graphics card :-\

Jeeeeeeeez …

I gave him a real earful at this point:

  1. Your build times are *far* too long … 1 month to install parts into a laptop? That should be 1 *day* …
  2. Your machines are uncompetitive. The Sony Vaio Z series is a 13” laptop that is just as well speced as the Studio 14 and is *cheaper* … and Sony ships in a week …
  3. Your web site sucks if it lets me pick illegal options …
  4. Your process for handling issues during the build is *terrible* … you don’t *cancel* a 1 month build cycle near the middle … that’s *insane* … you interrupt it and then *resume* it where you left off … that adds 1 day, not a week and a half.

I am having trouble thinking of Dell as a serious company nowadays … the above mentioned issues all impacted me on this order and lost them a customer. Probably forever as there are lots of brands out there. When you look at their lineup, one thing that strikes you is the mish mash they have made of the configurations … they make no sense together. Some offer one version of Office, others offer a different version. Some offer modern processors, others look like they were designed 5 years ago.

Dell … get it together.

F300EXR – Revier Part 14 – A walk in the woods with the new camera …

As mentioned in Part 13, I exchanged my camera yesterday for one with a better lens. Although this one is still not perfect, I have found that the corners are quite acceptable now at 24mm. A slight blur top left, but otherwise pretty good. And the blur top left is not dramatic the way the blur top right was on the previous cam.

So I took it for a brief walk in the woods today to see if the corners are acceptable in practical use … and landscape in the woods is a pretty common usage pattern for a compact. I shot this lovely path at at 24mm and then at 31mm … both look good to me. A bit of processing for exposure and tone, but none for sharpening (except output sharpening after downsizing.) You can follow the leaves right into the corners at the bottom … and the corner bottom left is currently the worst one on the cam.

DSCF0049_24mm[1]

DSCF0051_31mm[1]

A little way up, I found a scene that looked kind of nice, so I set the tripod to record two images, one at f/4.4 and one at f/13. This will reveal two things … that the corners improve slightly with the extra aperture, and that the middle softens slightly … this is either from slight tripod shake (although that would have affected the corner crop too) or it is diffraction beginning to set in (my guess.)

DSCF0056_crops[1]

DSCF0058_crops[1]

Now let’s see them without the crops obscuring the show …

f/4.4

DSCF0056_pmode_f4_4[1] f/13

DSCF0058_mmode_f13[1]

Both are perfectly acceptable …

By the way … the light was very soft here so I switched to Velvia for the last two. I shoot Velvia for the rest of the series, except where I compare them near the end.

Here is a spectacular bit of fungus … the colors are amazing … so much so that I had to juice them even further.

DSCF0060_fungus[1]

And a lovely maple leaf (Oh Canada!) hanging in the breeze … shot at the full 360mm … makes for some very decent looking backgrounds …

DSCF0061_maple_leaf[1]

The following image is a blend of two images. At 360mm from a close distance (4 feet above on tripod), the leaf un the upper left quadrant cannot be in focus at the same time as the log and ground. I happened to shoot it both ways and did not like either version … but painting the leaves in over top fixed the problem perfectly.

DSCF0067_blend[1]

Another great looking bit of fungus. The colors again are spiked because the image looks so nice when they are drawn out.

DSCF0070_fungus[1] If you clicked through, you probably noticed how sharp and detailed this image is. It’s pretty impressive. So here is an 800x800 crop from the middle … you will need to click on it to see it pixel for pixel … and no one can deny the inherent sharpness of this lens.

DSCF0070_crop[1]

I thought I’d revert back to Provia to improve the captures … I am not a big fan of black shadows. This was a good opportunity to show the difference with nothing else changed.

DSCF0072_film_simulation[1] You will definitely want to click through on that one …

And one last image … a tree with a small plant growing in its bark. Weird. The tree almost looks burnt and the background looks pretty smooth, despite being a tad busy …

DSCF0079_plant_on_tree[1] My verdict so far is that I am pleased with this copy of the F300EXR.

F300EXR – Review Part 13 – Swapped for a new cam …

Well, you knew it had to come … lucky part 13. And in this part, we see the first images from the new camera. I went to Henrys and did my best to find a better copy. I checked two cams and the second was much, much better than the others. So … here we are …

Interestingly, this cam still has plenty of CA in the corners and is somewhat soft at 24mm. But then, this is a real torture test and even my F80 does not fair well … and it has lovely corners in practice.

I have not shown this before, but my setup for these tests relies on an old, junky tripod that I can lean against the window sill and stabelize by extending the back leg further than the others. This works surprisingly well, although it does move quite a bit when I am handling the camera to change settings.

DSCF0044_test_setup[1]

So let’s see the crops. Here are the crops as shot on the left side, and defringed on the right. This process takes me only seconds, so it is not onerous when it shows up. But it would be nice of Fuji could make a perfect lens one of these days :-)

corner_crops_new_f300[1]

Those of you who hate these long zoom EXR cameras are not going to change your minds :-) But believe it or not, I consider these results in this torture test to be acceptable at 24mm. The center is fantastic, and the slight blur and fairly strong CA is only visible in the deep corners … the rest of the image is far better than on my first copy, where only the center of the image was protected from blur and CA …

But I cannot actually say for sure that this is acceptable until I blog my trip into the woods with this camera. Look for that shortly.

The two images look like this processed:

As shot:

DSCF0034_as_shot_24mm[1]

This test makes blur and CA blindingly obvious … and it really is obvious in this shot. But … it is again isolated to the deep corners, so I hold out hope that it will not heavily damage my images at this focal length. We shall see …

Here is the defringed version …

DSCF0034_defringed_24mm[1] That’s getting pretty close … the blur is still there but is *much* less obtrusive once the CA has been removed. This image is acceptable to me … we’ll see how it fares in real life …

Tuesday, September 28, 2010

Guitar Mind F*ck

Three guitars battle it out … but which one does not belong? This must be seen to be believed! I’d like to thank my friend Tony for posting this to Facebook …

Canada appears to have first human-powered Ornithopter!

For those who date themselves by professing a love for the Dune Trilogy by Frank Herbert, the concept of Ornithopter is intimately familiar. An airplane that flaps its wings like a bird. Simple as that.

But imagine if you could stay aloft with only the power of your legs?

F300EXR – Review Part 12 – 3D

Ok, this is a fluff piece. I enjoy fooling around with 3D, so I thought I’d shoot a bit of 3D magic with a setup I saw on one of the forums. Probably FTF. Sorry for forgetting who showed it.

I bought a set of Adorama macro rails pretty cheaply on fredmiranda.com a couple of years ago. I don;t use them often, but they are indispensible for close macro work and product shooting. With a powerful lens set to closest focus, it is trivial to dial the whole rig back and forth until perfect focus is achieved. With magnified live view, there is no excuse for a lack of precision.

Of course, this shoot is a very different animal … place the F300EXR in the rig by attaching a spare lens plate to its bottom, then align it to where the shot should go (I got lazy and did a half-assed job) and finally shoot. To shoot the two images needed for stereo viewing, you shoot one and then dial the image to the right by a few centimeters and then shoot again.

Here is that setup …

DSCF4851_setup[1]

You can see it moved around two centimeters … just enough to focus on one side of the tiny statue and then on the other.

Next, you take the two images and place them on the same image side by side (don’t accidentally swap left for right here) and leave a black band in between of something like a centimeter or less. Enough so crossing the eyes is comfortable. Here is that image and I find it really easy to lock in a 3D image.

DSCF0404_f300_3D[1]Please click through and view the larger version :-)

Edit: Bill’s been busting my chops down below in the comments about reversing the order of the images … left on the right and right on the left. Well, I should not have doubted him on that … his research skills exceed my own and he is dead right. Here is that version, and the 3D effect is much more pronounced and easier to achieve and hold. Sheesh :-)

DSCF0404_f300_3D_reversedThanks Bill.

And finally … a trick I learned somewhere on the forums. I can create an animated gif with the main subject aligned, then swap frames quickly and your brain will see the image in 3D. I showed it before with a real hacked attempt … this time it aligns very closely. Way cool … again, click through for the larger one.

DSCF0404_3d_animation[1] Try it … it’s a hoot :-)

Monday, September 27, 2010

F300EXR – Review Part 11 – Deep dive into the corners and the effects of the 3rd stop of aperture

I’ve been wanting to set up the tripod near my window screen to get a proper juxtaposition of 1st and 3rd stops (the two real apertures) against the first 8 clicks of zoom, or thereabouts. Today, I spent a few minutes setting up the tripod against the wall with the back leg longer and the center column fully extended. This gave me a nice position in the middle of the window where the top corners would see uniform grey sky. Exactly what is needed to test the chromatic aberration and blur in the top right corner, which has been previously proven to be the worst one on this particular sample of the F300EXR.

What I found is that 24mm seemed to change very little … I suspect that this is because the blur is bad enough to swamp the small improvement at the 3rd aperture (f/10 in this case.) On the other hand, as I began zooming in, there was noticeable – if fairly small – improvement at each zoom range. It manifests more and more as an improvement in the thickness of the dark portion of the screen in relation to the fringes of color on each side. When looking at a processed image, the difference if fairly obvious and thus, I think it is safe to suggest that landscapers consider using the 3rd stop from a tripod if they want the best possible images. Even hand held will be ok on very bright days, but shoot several to put statistics in your corner … IS being purely statistical where sharpness probabilities are concerned.

The images I shot were at 800 ISO in order to get decent shutter speeds at longer zoom … the lowest dropping 1/100s. There is some grain in the images, but the effects of noise reduction (NR) on edges is actually clarified by the risk of higher NR.

The first set of crops are the top right corner at both apertures and at many zoom ranges. Generally one click inward for each sample. As I reached the last sample, I was forced to shift the tripod back a few centimeters as I finally ran out of focus range.

top_right_corner[1]

And now a selection from inside the image, showing a subtle improvement in sharpness of the very low contrast, almost gossamer, cob-webs (nothing like a macro image of a screen to show you where you forget to vacuum :-) … you can actually see the NR attack the threads in the bottom middle panel on the left whereas it leaves it alone in the bottom middle panel on the right. So what is happening? My guess is that the thread is just that much sharper, thus escaping being classified as noise. The edge it thin with the EXR cameras, and it takes very little extra light to make an image look a whole lot sharper. Something to remember …

Also note that again I prove that there is a real aperture at the 3rd stop … the window in the middle of the frame is almost faded out completely on the left, and on the right it regains its distinct rectangular shape. Go Fuji!

away_from_corners[1] Click through on that one to see a bit better what is going on … especially that NR attack.

So … two things you can do if you get an F300EXR with one or two rather soft corners …

  1. Zoom in a wee bit
  2. Use the 3rd stop

Of course, the second one may necessitate the use of a tripod. Sorry about that …

Edit: I noticed an area where the difference is simply stunning … just inward from the top right and down a bit … so perhaps 10% in on a diagonal. This will blow your mind …

animated_blur[1]

Sunday, September 26, 2010

Important Tips for your Photography

This author presumes that quite a few of the readers of this blog are interested in getting better at certain aspects of photography. While I have never claimed to be good at it, I am certainly much better at it than I was a few years ago. I have decided that, in lieu of a web site dedicated to teaching with a proper structure for that, I will maintain a tips post and update it as I see fit. You will find it linked in the recommended posts at the top of my blog.

Tip 1 – Calibrate Your Monitor

The odds are very good that you are not calibrating your monitor ... most people do not. But this leaves you at the mercy of the manufacturer as to how accurately you see colors in images. Especially poor from the factory is the white and black points, so you miss all the subtlety in tones. Seriously, consider it. See my detailed article on the subject here.

I use the Huey Pro, as it handles all your monitors and has very good software in my experience. My results were amazing.

Tip 2 – Please Use a Quality Panel for Processing

Now that you have decided to calibrate your monitor (I hope), consider getting yourself a decent monitor. The traditional gaming and general purpose panel is cheap and is based on TN technology. Crap. For photography, you need something with a very wide field of view and the ability to generate true 32-bit color or better. I use an NEC MVA panel, a Dell 24" 16:10 IPS panel, and a Dell 23" IPS panel and they all slaughter my TN panels. Apple makes the amazing cinema line of monitors that use the best technology -- S-IPS, but my 24" Dell uses H-IPS I believe and it is no slouch.

You can of course get other brands with cheaper S-IPS, E-IPS, PVA or MVA technology. All will be a massive cut above TN technology for editing, especially when calibrated. Here is one panel from Asus for an amazingly good price. There are many such deals on Amazon and there is little excuse to continue to use a monitor that cannot perform well against the test pages at http://www.lagom.nl/lcd-test/

Tip 3 – Shooting Concerts

A lot of people like to shoot images at concerts. I find it very rewarding and I get to relive the concert any time I choose. But most people start by shooting in Auto mode and that makes a hash of most concerts. The shutter speeds are wrong, the ISO is usually wrong, and the exposure in general is wrong. This leads to burnt out highlights, blur, severe grain, or any combination of all three. I certainly had my share of these problems at first. But I’ve developed a technique that tends to work fairly well.

If your camera allows it, set the shutter speed to 1/100s in shutter-priority mode or manual mode. This is the starting point for shutter and you will change it as a last resort. 1/100s generally works well enough to freeze motion on stage while still allowing a lot of light into the image.

Make sure your aperture is wide open if you are shooting manual. You need the most light you can get.

Start with your ISO ar 400 or 800, and your compensation at –1ev (if in S-Prio, in manual you do not need to use compensation as it will only confuse matters.) Low ISO will show up as fairly dark images on the LCD, but this is much better than burnt or blurred images. A somewhat darker image can be lifted, while a too-bright or blurred image is fatally wounded. Remember that … blur is forever, burn out is forever, but grain and somewhat dark can be fixed.

To fix a slightly too dark image, lift its tones in an editor. I did this for an image of Andy Maize of the Skydiggers, this was shot at 800 ISO with an F70EXR in an excruciatingly dark venue – the rest of the images there were shot at 3200 ISO.

DSCF2379_sky[1]

I also did it for this 500 ISO shot of Bruce Cockburn with the D300:

DSC8307_bruce_cockburn[1]

Obviously, the low ISO cuts out a lot of ambient (stage) light, so you will end up with the performer as the subject and little else in the image. But these were dark venues, so there was little to see anyway. Raising the ISO can raise the amount of light you get and reduce the amount of editing, but it also risks burn out and severe grain, so experiment with your camera at different venues to see how you like to shoot.

Here is a 3200 ISO shot at the same dark venue as the shot of Andy Maize above, This is Billy Bragg:

DSCF2454_billy[1] That’s really not too bad for a compact cam at 3200 ISO. And here is why … the subject was actually fairly well lit, and the extra light helped increase contrast and keep the Fuji noise reduction algorithms at bay. This matters a lot … shoot in low light and noise reduction smooths everything over … shoot at a concert and the subject tends to look decent even at higher ISOs. Use that to your advantage when you need to, but try to keep ISO low anyway.

When the stage is extremely well lit, you can shoot at 400 or 800 ISO and get superb crispness. This shot from the F70EXR at 800 ISO of Mama Mia! is a case in point:

DSCF3089_sisters[1]

Another example from a Rascal Flatts concert:

DSCF3324_rascals[1]The F80EXR or F300EXR should do just as well as they are terrific at 800 ISO. Many other cams are as well, although they do not have this reach. I doubt that the ZS series from Panasonic can cut it at 800 … but maybe. And this shot could probably be replicated at 400 ISO … at 1/50s perhaps. But more risk of blur ensues.

So go forth and enjoy concert shooting. But control your camera and learn how to tweak the exposures in post processing. It will make a *huge* difference for you.

Edit: In the comments, Adrian mentions that the #1 tip should be to turn your flash off. Yes, definitely. Your flash is only useful if you are standing within 10 feet or so of the stage, and even then it does not add enough light to be worth burning down your battery. What it does though is light up the crowd near you, showing their hair etc, and thus destroying what might have been a good shot. You want the heads in front of you to be black silhouettes … keep the flash off! Thanks Adrian.

Tip 4 – Read!

I apologize if this tip is somehow patronizing, but the majority of people do not read enough about their hobby from what I have read (the irony of that statement is not lost on me :-)

There are experts out there who have put together superb treatments of one or more topics and you should absolutely take their knowledge and blend it with your own. If you read only one book, I suggest it be John Shaw’s Field Guide to Nature Photography. I find that book an excellent introduction to the basics … especially where metering and exposure are concerned.

Read my more thorough treatment of my favorite books here.

Saturday, September 25, 2010

Fuji X100 at Photokina 2010 – Forums go Wild! – Camera evolution changed forever!

The evolution of the camera has come full circle and there is change in the wind, apparently thanks to a modest yet expensive compact camera from Fuji … the soon to be crowned X100 if its many (instant) fans have anything to say about it.

I feel that this may permanently shift the direction in which manufacturers go from here on. No longer satisfied to baffle us with so many complex features that only get in the way of the essence of photography, this new thrust will more and more connect the photographer directly to the instrument. This will truly embody the notion of the “camera just getting out of the way” to allow the capture of the perfect moment.

I can’t wait to see the response of the big guns … after all, they have been tortured by the appearance of many relatively compact cameras with lots of features and interchangeable lenses with the fairly small 4/3” sensor. Now, Fuji raises the stakes by increases sensor size yet reducing control … if Canikon were worried before, they must be expelling bricks with the X100 on the horizon.

I see it playing out rapidly … perhaps over two or three years … simplification being the hallmark of this new era, I think we can look to the past as our guideline. I foresee a wonderful renaissance for photography as a hobby for everyone …

camera_evolution[1]You heard it here first …

Thursday, September 23, 2010

F300EXR – Review Part 10 – Check Corners from Close Distance

I shot my bedroom window screen in macro mode to see how the corners look close in. The issue is not so much blur as it is unbelievable amounts of chromatic aberration. As shown here:

DSCF0226_screen[1]

Click through to see it. Here are some crops, also better viewed at the full 800px size.

DSCF0226_screen_crops[1]

I do need to exchange this one.

I shot this image of Karen at 80mm effective with direct flash inside in very low light. The F300 at 800 ISO is pretty darned good in my opinion.

DSCF0264_karen[1]

F300EXR – Review Part 9 – Depth of Field Explored

Now *this* was interesting … I had promised to devise a test to determine if Fuji added two ND filters for the 3-level aperture in the F300 or if perhaps they had mercy on us and added a real aperture somewhere in there.

So I set up the tripod(s) about 5 feet from my lovely Canon Lens coffee cup (thanks Nick) with a door knob about 6 to 8 feet behind it. This ration of 5:8 in subject / background distance at close quarters will give pretty nice blur at 270mm, even for these small sensors.

As a reference for DOF, I also shot the D700 in this test with the 70-300VR on tripod but from an extra foot back as I was unable to get focus that close (the EXR cams were in macro mode.)

Now … sensor size differences predict a pretty massive difference in depth of field. The crop factor for this 1/2” sensor with this lens is 5.4, which means that f/5.6 should have DOF equivalent to about f/30 on the full frame sensor. (5.4 * 5.6 = 30)

So here is how the D700 shot the scene at f/5.6.

DSC_6891_d700_5_6[1]

And here is the F300

DSCF0310_f300_5_6[1] Yeah … that’s pretty different :-)

To verify that the F300 has not changed from the F70, here it is

DSCF4847_f70_5_6[1] Same DOF. The entire test was shot at 400ISO in order to get shutter speeds that were comfortable for the F series cams. They max out at 8 seconds.

So, without further ado, I show you the complete set of crops. And yes, you absolutely *must* click through, lest someone think I expect you to judge from the 400px blog version :-)

DOF_test_exr_d700_crops[1] Some observations from this test:

  1. The F300 has a real aperture at f/16!!! DOF increases.
  2. The D700 retains smoother bokeh past the point at which I expected it to match. But it gets close at f/40, which is close enough to the calculations when you consider how approximate the formula is.
  3. There is vignetting wide open on the F70 and F80, this has been known for a year and is confirmed in this test.
  4. There is *no vignetting* wide open on the F300 at 270mm. Well done Fuji!

Wednesday, September 22, 2010

F300EXR – Review Part 8 – Aperture Range Across the Zoom

I noticed something rather peculiar … I was seeing very high apertures in the middle of the zoom range and a fairly low aperture – f/5.3 – at the end of the zoom range. In fact, the F70 and F80 are both at f/5.6 at 10x, which for them is 270mm equivalent. The F300, on the other hand, is at f/6.2 at 10x, 1/3 stops slower, and then falls to f/5.3 at 15x … 1/6 stop faster than the F70 and F80 at 10x.

It’s all very peculiar … watch the video here to see this in action.

The strange audio whistle you hear is not in the file I uploaded … that was added courtesy YouTube compression :-(

Edit: The next morning, the audio whistle is gone. So YouTube was still processing it in the background while I was watching it. Cool :-)

Tuesday, September 21, 2010

F300EXR – Review Part 7 – Check Those Corners

I stopped on the way home from work today at a location where I could shoot the F70 against the F300 (only two pockets in my coat) … from previous tests, the F80 is actually my best cam for corners, the F70 not far behind but a bit soft. The F300 is clearly behind the F70, but of course that’s at 24mm.

What I will do when I get a chance is find a nice brick wall and try to get a clear idea of the amount of blur from each of these cams. Meanwhile, here are my results from today.

F70EXR

DSCF4818_f70_m43[1]

Those corners are straight from the cam and untouched by any processing. Top left and right are slightly blurred, bottoms are both fine by me. I have been shooting the cam for a long time and really don;t mind it at full wide, so I don’t worry too much about these results on a megazoom.

F300EXR

DSCF0181_f300exr_m43[1]Top left is more blurred than F70, but tolerable. Top right is pretty bad, enough that I should take it back and get it fixed. Bottom left and right are acceptable.

So we’ll see … I bought three years of Internantional “help” from Henry’s so geting it fixed is a piece of cake. After this round of tests is completed, I’ll pop it in for warranty or replacement.

Now … here are the two images fully processed.

F70

DSCF4818_f70_m43_pp[1]F300

DSCF0181_f300exr_m43_pp[1]  At web sizes, the top right corner is noticeable, but not necessarily fatal to my images. Remember that I use compacts for general shooting and concerts. The corners just don’t matter all that much to me as I have the Tamron 28-75 2.8 on the D700 that I can stop down to f/8 to get perfectly clean shots.

Don’t bother comparing the processed images for color … I mucked with saturation etc and warmed them with photo filter. I prefer the tone of the F70, but the F300 could have been made to look the same with more time put in.

Next I shot some bull rushes nearby. Again we see the difference in magnification between 270mm and 360mm. We also see the difference between 100iso and 400iso … I’m not 100% sure how the F300 went to 400iso so early, but I was shooting L4:3 with each cam and as we know the EXR cams will either set DR100 or ISO400 if coming from ISO100 / DR400 in M4:3 mode and it is not always obvious why one is chosen instead of the other … this to me is a huge flaw in Fuji’s algorithms, but so be it.

Note also that it took me *4 tries* for each cam to get the final crisp shot. It was windy today and the rushes were blowing around. The F300’s fancy phase detect focus again did nothing to help, as with the moon shots yesterday.

F70 processed and then unprocessed with crops

DSCF4822_f70_bullrush[1]

DSCF4822_f70exr_m43_crops[1]

Now that’s crisp. ISO 100 helps a lot.

F300 processed then unprocessed crops

DSCF0187_f300exr_m43_400iso_pp[1] DSCF0187_f300exr_m43_400iso[1]

Not bad, but the F70 wins this battle. Still, 400iso versus 100iso is hardly fair, especially considering the F70 has lower pixel density. Also, the longer lens makes the plane of focus considerably thinner … the top crop looks like it is starting to fall outside the crispest part of the focus.

I like both, but the processed F300 stands out for me as better because of the increase in subject isolation.

I then turned my attention to a daisy on the ground. Standing next to it, I shot from fairly close and got this shot with the F70:

DSCF4825_f70exr_daisy_100iso[1]

Pretty crisp in my opinion. Nicely held highlights on the petals.

Now the F300 from closer in, just because it can:

DSCF0190_f300exr_daisy_400iso_shadow[1]

Also a very nice shot, but note the shadow across the right edge of the petals. Oops … just because you can get really close, does not mean that you should get really close. It blocks the light or creates reflections, depending on the subject. This is something to remember when touting the super-macro setting of small sensor cams … getting that close blocks out the light and makes the mode useless at times in the field. Macros always look better with an achromat at full zoom.

Here’s the better one with the angle slightly changed and the distance increased a tad …

DSCF0191_f300exr_daisy_400iso[1]

A very crisp looking shot. I’m quite pleased.

Now. my final shot for this series … a shot of a strange track along the side of the road with trucks in the distance.

F70 @ 28mm

DSCF4827_f70exr_28mm_400iso[1] F300 @ 24mm

DSCF0196_f300exr_24mm_400iso[1]

The corners continue to look a bit blurred. But overall I think the image competes.

What to conclude … 24-360 is a bit much. The corners are blurry at full wide, which is where you expect that. Same issue as with the HS10. Pretty difficult to design something like this.

If you are a pixel peeper, you might want to continue shooting your F200EXR and revel in the better corners … a product of having much less reach. But I would not necessarily shy away from the F300EXR because of this level of blur. You can always shoot M3":2 instead of M4:3 … I’ll set up a test specifically to look at the different formats at 24mm and the to shoot at each zoom incremement until we pass about 30mm … also to check on the effect (if any) of stopping down. With three apertures now, the middle one may actually make a sharpness difference as there may actually be an aperture involved.

The F300EXR continues to feel good in the hands and while shooting.